PROJECT SUMMARY/ABSTRACT Anxiety is a debilitating symptom of most psychiatric disorders including PTSD, major depression, schizophrenia, autism and addiction. Treatment of anxiety is mostly limited to benzodiazepines, which have abuse potential and produce multiple cognitive and behavioral side effects, including increased propensity to develop dementia. Design of alternative treatments or prevention strategies is contingent upon a better understanding of the neuronal basis of anxiety. While animal studies have so far informed us about the functional neuroanatomy of fear and anxiety, the negative impact of real-life anxiety extends beyond aversive feelings and involves disruptions in ongoing goal-directed behaviors. For example, anxiety is associated with deficits in flexible control of reward-driven actions and expression of motivated behavior when it is subject to the risk of an aversive outcome. The neural basis of these behavioral deficits is largely unknown. Thus, the overarching research question that drives the experimental aims of this application is: How does anxiety affect neuronal encoding of goal-directed behaviors? To address this question, a major challenge of the experimental approach is to create a background state of anxiety that does not prohibit animals to perform goal-oriented and rewarded tasks during electrophysiological recordings from multiple regions. With this in mind, we propose to use two complementary experimental models of anxiety in combination with innovative and clinically relevant behavioral tasks while measuring dynamic coordination between neurons of two regions implicated in reward processing and flexible control of behavior: ventral tegmental area (VTA) and dorsomedial prefrontal cortex (dmPFC). Specific aims are designed based on a computational-style model with assumptions that are supported by preliminary data: (1) behavioral differences (between control and anxiety states) are specific to conditions that involve action selection under conflict or the risk of an aversive outcome; (2) differences (between control and anxiety states) in neuronal activity are observed during conflict/aversive outcomes and involve diminished recruitment of action encoding neurons in dmPFC and disrupted coordination between dmPFC neural activity and VTA dopamine neurons. This approach is novel and significant because a neurocomputational understanding of aberrant neural activity relevant to symptoms such as anxiety can help identify biological markers and clinical measures that delineate etiology and physiology of those symptoms.